The ECtHR Finds Macedonia Responsible in Connection with Torture by the CIA, but on What Basis?The ECtHR Finds Macedonia Responsible in Connection with Torture by the CIA, but on What Basis?

 Posted by: Nollkaemper André

Cross-posted from EJIL: Talk!

On 13 December 2012, the European Court of Human Rights (‘the Court’) found that the Former Yugoslav Republic of Macedonia (‘Macedonia’) was responsible in connection with the ill-treatment and torture of Khaled El-Masri. The judgment adds a further chapter to the Court’s rich case law on situations where a state party is held responsible in connection with the (wrongful) acts of another state.

El-Masri, a Lebanese-born German national, alleged that in the period from 31 December 2003 to 29 May 2004 he had been subjected to a secret rendition operation, in which agents of Macedonia had arrested him, held him incommunicado, questioned and ill-treated him, and handed him over at Skopje Airport to CIA agents who then transferred him to Afghanistan, where he had been detained and ill-treated for over four months.

No one who reads the facts of the case will argue with the Court’s conclusion that Macedonia had to bear international responsibility. The question is on what grounds one can base this conclusion.

The approach chosen by the Court may surprise many international lawyers. Influenced by decades of work of the International Law Commission (‘ILC’), their approach would be a combination of attribution of conduct on the one hand and the breach of an international obligation, on the other: Macedonia then would be responsible for handing over El-Masri to the CIA, in the face of risk (if not certainty) that he would be ill-treated and tortured. They would not normally say that the act of ill-treatment at the hands of the CIA itself is attributed to Macedonia, but limit Macedonia’s responsibility to its own wrongful conduct. This distinction may seem a legal nicety, but it may have practical relevance (for questions of evidence and reparation) and also reflects that what is essentially a sovereignty-based consideration: it should not easily be presumed that a state is responsible for acts committed by another subject of international law.

The Court takes a somewhat different approach. But it is quite difficult to figure out exactly what this approach is. While the fact that the Court does not feel compelled to follow the ILC’s conceptual straightjacket is in many respects refreshing, its own line is at times somewhat inconsistent and confusing. For one thing, it is difficult to see why the Court uses interchangeably the terms ‘attribution’ and ‘imputation’ – one may guess that the Court uses the latter when it seeks to leave aside the ILC’s approach, but it would be nice if the Court would not invite us to speculate.

A more substantive point is that, without clear rationale and justification, the Court intertwines the concepts of attribution of conduct and attribution of responsibility. For instance, the Court found that the responsibility of Macedonia was engaged with regard to the applicant’s transfer into the custody of the United States authorities and eventually his transfer to Afghanistan, despite the existence of a real risk that he would be subjected to further treatment contrary to Article 3 of the European Convention on Human Rights (‘ECHR’) (para. 223). This follows the Soering case law, and fits in traditional doctrine: Macedonia would only be  responsible under Article 3 for its own conduct and not for the torture in Afghanistan itself. But it is then hard to understand why the Court speaks in this context of attribution of responsibility (para. 215), rather than attribution of conduct.

Nonetheless, amidst all this confusion, an interesting thought emerges. In regard to two allegations, the Court concluded that Macedonia was not (only) responsible for the act of handing over El-Masri, but that it was responsible for conduct that clearly was not its own. Macedonia had handed over El-Masri to a CIA rendition team at Skopje Airport, where he was ill-treated before being sent off to Afghanistan. The Court found Macedonia ‘responsible for the ill-treatment to which the applicant was subjected at Skopje Airport by CIA agents’. And in respect to the detention in Kabul, once the CIA had flown EL-Masri there, it found Macedonia to be responsible for a violation of Article 5 during the entire period of his captivity in Kabul (para. 240), and imputed the detention (an act committed by the CIA) as such to Macedonia (see also para. 235).

This approach of making a state responsible for acts that are not its own may be explained by the scope and contents of the particular obligations of states under the ECHR. That is particularly clear for the events at Skopje airport. The Court could have approached this as a Soering type situation, and could have held Macedonia responsible for handing over a person to a foreign state in the face of a risk of torture. However, the difference was that the CIA mistreated El-Masri at the airport itself, and this was therefore not just a question of handing someone over in the face that he would be mistreated in another state. The Court thus had to take a different approach and stated in para. 206 that:

the acts complained of were carried out in the presence of officials of the respondent State and within its jurisdiction. Consequently, the respondent State must be regarded as responsible under the Convention for acts performed by foreign officials on its territory with the acquiescence or connivance of its authorities.

The Court’s finding that Macedonia was responsible under the Convention for the acts performed by foreign officials on its territory with the acquiescence, or connivance of its authorities, reminds us that there is a world outside the conceptual framework established by the ILC. The Court did not attempt to explain its approach in ‘ILC terms’ – none of the four Articles of the Articles on State Responsibility (‘ASR’) that the Court referred to (Arts. 7, 14, 15 or 16) is relevant in this context. Driven by the general obligation of states to protect rights of persons under their jurisdiction, the Court’s approach seems to be Macedonia is responsible for the wrongful act of handing over El-Masri and that because of its acquiescence or connivance, this responsibility extends to El-Masri’s ill-treatment by the CIA.

This approach follows the Human Rights Committee, which in a 2006 case against Sweden determined that ‘a State party is responsible for acts of foreign officials exercising acts of sovereign authority on its territory, if such acts are performed with the consent or acquiescence of the State party’ (Mohammed Alzery v. Sweden, para. 11.6).

A striking aspect of the Court’s reasoning is that it equates the responsibility of a state vis-à-vis the conduct of another state (here the US), with the responsibility of a state vis-à-vis the acts of private persons. It refers, for instance to Ilascu v. Moldova and Russia, in which the Court said that ‘the acquiescence or connivance of the authorities of a Contracting State in the acts of private individuals which violate the Convention rights of other individuals within its jurisdiction may engage the State’s responsibility under the Convention’ (para. 318).

The Court’s equation of state responsibility vis-à-vis conduct of private actors on the one hand, and vis-à-vis foreign states on its territory, is logical from the perspective of the ECHR. Once the scope of states’ obligations to protect rights of persons within their jurisdiction was extended to obligations to secure, that third parties within their jurisdiction did not interfere with rights and freedoms of individuals, there is no compelling basis for distinguishing between private parties and states that acted within the territory of the state party.

The justification of the construction then lies in the combination of the (positive) obligations of states party under the Convention, and the fact that the conduct in question took place on its territory with its acquiescence or connivance, which in turn was incompatible with the positive obligations. The primary rules thus in a way incorporate questions that in the ILC texts are considered as freestanding secondary rules, just as the Court considers the question of facilitation of ill-treatment of El-Masri as a breach of positive obligations (e.g. para. 211), not in terms of a responsibility under Article 16 of the ILC Articles (even though the Court did list this as a relevant norm of international law).

Although all of this thus can be explained with the framework of the Convention, a reading of the judgment raises many questions concerning this approach. The statement that Macedonia was responsible under the Convention for acts performed by foreign officials on its territory is somewhat ambiguous. Since the Court did not go as far as attributing CIA conduct to Macedonia, this wording may be taken to suggest that Macedonia would be responsible without itself having committed a wrongful act. That of course is not what the Court intended to say, but a more solid explanation would help.

Somewhat mysterious is also that whereas Macedonia’s responsibility for handing over El-Masri to the CIA was based on the traditional Soering case law (Macedonia was responsible for the act of transferring El-Masri into the hands of the CIA in the face of risk of torture (para. 223), not for the torture in Kabul, just as in a Soering case, the extraditing state is not responsible for the death row or the death penalty that later occurs), it took a different line in regard to the responsibility of Macedonia for the illegal detention by the CIA of El-Masri in Kabul. Here the Court found Macedonia to be responsible for a violation of Article 5 during the entire period of his captivity (para. 240) and suggested that it imputed the detention (in the hands of the CIA) as such to Macedonia (see also para. 235).

This could be explained as extension of responsibility based on criteria of foreseeability and causation, but then it is not obvious why these factors were not applied in regard of the removal of the applicant to Kabul, where the Court only held Macedonia responsible for the removal, not for the torture in Kabul itself. All of this seems a connection of unfinished and not systemically developed thoughts.

Yet, in this midst of all this there is a thought that matters. For all its incoherence and lack of clarity, the Court’s language has a hint of normative power that the general law of responsibility lacks. The general law of responsibility by its conception of responsibility-based-on-wrongfulness, prefers determinations that one is responsible for the handing over of a person or for its inaction, not for the resulting torture itself.  In contrast, the Court’s approach may allow us to say that if a state hands over a person to another state in the knowledge that the person is tortured, and stands by when that torture happens, it bears responsibility for the torture itself.

The hint of a fresh approach to the law of responsibility deserves to be taken forward. There is no other international Court that has a similarly rich case-law on responsibility of states in connection to conduct of other states. It is about time that the Court starts clarifying the grounds of responsibility of states under the Convention, and in particular the grounds for extending such responsibility to acts that were committed by other states.

Syrian Refugees, Allocation of Responsibility and the Right to Choose: Beggars Can’t Be Choosers (or Can They?)

 Posted by: Swerissen Isabelle

On 11 December 2012, the United Nations High Commissioner for Refugees (UNHCR) announced that there are more than half a million registered Syrian refugees. That number is climbing by the day, as more refugees flood into neighboring countries. In Turkey, for example, 136,319 refugees are living in 14 government-run camps. From there, many of them try to make it into Greece, the gateway to Europe, only to be returned to Turkey. This practice of sending refugees back is currently taking place covertly, but the European Union (EU) is in the process of negotiating a readmission agreement with Turkey.

Through the conclusion of readmission agreements between destination countries and transit countries, the latter undertake to readmit third country nationals, including refugees, into their territory. As such, the agreements are part of a wider phenomenon which I refer to as allocation of responsibility for processing a claim for protection. Such allocation of responsibility is premised on the idea that refugees do not get to choose their preferred destination. Refugees are the world’s supplicants – and ‘beggars can’t be choosers’. Whether refugees are indeed legally unable to be choosers is considered in this blog post. How much legal freedom, if any, do refugees have in choosing the country that will decide their claim for protection? And if refugees do have some element(s) of choice, how should this be reflected in the terms of the future EU-Turkey Readmission Agreement?

The right to choose has been the subject of some modest debate. Two diametrically opposed positions can be identified. James C. Hathaway, on one end of the spectrum, observes that states have the sovereign discretion to decide who obtains admission, subject only to international obligations. The quid pro quo of this discretion, he argues, is that refugees get to decide where to take their chances in applying for protection.[1] Kay Hailbronner, on the other end of the spectrum, opines that it is not the prerogative of refugees to pick and choose where to go. In his view, refugees are supposed to apply for protection in the first safe country they reach.[2] An adequate reading of the law may lie somewhere between those extremes.

The vantage point for an examination and analysis of the law is Article 31 of the Convention Relating to the Status of Refugees (1951 Convention). This article first provides that “States shall not impose penalties, on account of illegal entry or presence” on refugees “coming directly from a territory where their life or freedom was”. The criterion of ‘coming directly’ has been used to support the view that the 1951 Convention does not allow for choice on the part of refugees. However, Article 31 continues by recognizing that refugees may  have “good cause for their illegal entry or presence”. The criterion of ‘good cause’ is, in the words of Guy S. Goodwin Gill, “flexible enough to allow elements of individual cases to be taken into account”.[3] This points towards the existence of some degree of choice for refugees.

In order to solve this ambiguity, the precise meaning of the terms ‘coming directly’ and ‘good cause’ should, in line with the Vienna Convention on the Law of Treaties, be interpreted in the context and in the light of the object and purpose of the 1951 Convention. The primary aim of the 1951 Convention is to extend the protection of the international community to refugees and, as specified in the Preamble, to assure to “refugees the widest possible exercise of … fundamental rights and freedoms”. When applied to Article 31, this indicates that the provision was first and foremost meant to ensure that states will not refuse admission to refugees on the pretext that they had entered their territory without authorization. After all, this would have undermined the raison d’être of the 1951 Convention.

The main purpose of Article 31 is thus to prevent the imposition of penalties. This interpretation still leaves open the question whether states are permitted to send refugees back to a transit country, without refugees having any choice in the matter. However, in R v. Uxbridge Magistrates Court and Another, Ex parte Adimi, which covers one of the most thorough examinations of the scope of Article 31, this question was decisively answered by Lord Justice Brown. He rejects the argument that Article 31 allows refugees no choice as to where they should claim asylum. Instead, he contends that ‘some element of choice’ is indeed available to refugees and writes that:

[A]ny merely short term stopover en route to such intended sanctuary cannot forfeit the protection of the Article, and that the main touchstones by which exclusion from protection should be judged are the length of stay in the intermediate country, the reasons for delaying there (even a substantial delay in an unsafe third country would be reasonable were the time spent trying to acquire the means of travelling on), and whether or not the refugee sought or found there protection de jure or de facto from the persecution they were fleeing.

The view taken by Lord Justice Brown in the Adimi case is supported by Justice Newman who, in the same case, considers that “as a result of the distinctive and differing state responses to requests for asylum … there exists a rational basis for exercising choice where to seek asylum”. In addition, in Charles Kofi Owusu Ansah v. Minister of Employment and Immigration, the Canadian Federal Court declared that, in certain circumstances, refugees may not need to apply for protection in transit countries before reaching Canada. The Court specified that language, family ties and cultural bonds are credible and sufficient reasons to account for a refugee’s decision to ask for protection in a country other than the one of first arrival.[4]

The case law on this topic, of which this blog post does not give an exhaustive account,[5] bears strong resemblance to the approach of the Executive Committee of the High Commissioner’s Programme (ExCom). ExCom is a body comprised of government representatives from 87 states to provide guidance and ensure consistency in applying the terms of the 1951 Convention (and its accompanying Protocol). To that effect, it periodically issues ‘Conclusions’ which are part of the ‘soft law’ body of international refugee law. In one of its early conclusions, ExCom gives some support to the idea of a limited choice for refugees. According to Conclusion No. 15 (1979), the existence of “a connection or close links with another State” should be observed when allocating the responsibility for determining a claim for protection.

The EU has adopted a near identical standard in the Procedures Directive. The Procedures Directive aims to harmonize procedural guarantees given during the asylum procedure and, in doing so, confirms certain basic procedural guarantees. However, EU member states may decide not to give (access to an) asylum (procedure) to refugees on the assumption that they can find protection in a ‘safe third country’ through which they transited prior to coming to the EU. This situation is dealt with under Article 27 of the Procedures Directive, which stipulates that, before member states may apply the safe third country concept, the competent authorities must first be satisfied that there is “a connection [with] the third country concerned on the basis of which it would be reasonable for that person to go to that country”.

These and other references in international legal materials to the connection that a refugee may have with a particular country point towards the existence of a limited right to choose for refugees. Refugees may have several reasons for preferring to apply for protection in one state over another. While some of these reasons have gained selective acknowledgment, such as the presence of kin or friends and language or cultural affinity, it remains to be determined which elements of choice fall under the legal freedom that refugees have in this respect. There is a chance for the future EU-Turkey Readmission Agreement to contribute to the development of the law by elaborating on the exact scope of a right to choose as the international community continues to look for a solution to the outflow of Syrian refugees.

 

---

[1] James C. Hathaway, ‘Why Refugee Law Still Matters’ (2007) 8 Melbourne Journal of International Law 89, 90-91.

[2] Kay Hailbronner, ‘The Concept of “Safe Country” and Expeditious Asylum Procedures: A Western European Perspective’ (1993) 5 International Journal of Refugee Law 31, 58-59.

[3] Guy S. Goodwin-Gill, ‘Article 31 of the 1951 Convention relating to the Status of Refugees: Non-penalization, Detention and Protection’ in Erika Feller, Volker Türk and Frances Nicholson (eds.) Refugee Protection in International Law: UNHCR’s Global Consultations on International Protection (Cambridge University Press: Cambridge, 2003) 185, 194.

[4] Nazaré Albuquerque Abell, ‘The Compatibility of Readmission Agreements with the 1951 Convention relating to the Status of Refugees’ (1999) 11 International Journal of Refugee Law 60, 79.

[5] See Goodwin-Gill (n 3); Gregor Noll, ‘Article 31: Refugees Unlawfully in the Country of Refuge’ in Andreas Zimmerman (ed.) The 1951 Convention Relating to the Status of Refugees and its 1967 Protocol: A Commentary (Oxford University Press: Oxford, 2011) 1243.